With the turn of the century, the world was veritably bombarded with new inventions in the IT sector, with computers advancing at light-speed. It is now literally impossible to imagine a world without computers. From Fugaku(the world’s fastest supercomputer at 415.5 petaFLOPS) to Jiuzhang(the world’s fastest quantum computer), the world of electronic brains extends far beyond the calculators we use for our Maths homework. Essentially, all computers are identical: you input data, it crunches some numbers and provides an output. However, the time has now come wherein computers are surpassing notions of how solutions can be derived.
Have we ever stopped to think about how Google Translate works? Granted, it’s not ENTIRELY accurate, but it is useful when we want to sound snappy or get the general context of foreign speech. However, Google Translate does not know a WORD of English or Spanish or French or Hawaiian. Google Translate sifts through a vast amount of information to determine how frequently one word in one language has been translated into a word in the other language. In this way, the program can make an accurate translation without actually learning either language. This is similar to John Searle’s Chinese Room thought experiment, wherein a person without an understanding of Chinese could, nevertheless, “translate” a Chinese sentence into English or any other language by using a set of instructions, or rules, that would substitute for understanding. This raises questions about what it means to understand anything at all, and whether understanding implies that something else is happening in the mind besides following rules. This enables computers to unravel problems we cannot ordinarily solve, even given a hundred years, such as the recently solved Schrödinger’s equation and the new computer algorithm predicting the orbits of planets, literally doing the unthinkable. Let it be noted that the process DOES NOT INVOLVE the equations we would use: computers don’t need Newton’s theories of gravity to predict the force between two objects: they are given a set of data, from which they gather a gist of the equation, or rather, observe the pattern and apply that to the given problem.
So there you have it: machine learning, the most advanced sort of learning on this planet, and the only sort of learning that doesn't involve “learning.”
Comments